Subscribe to Our Newsletter

Success! Now Check Your Email

To complete Subscribe, click the confirmation link in your inbox. If it doesn’t arrive within 3 minutes, check your spam folder.

Ok, Thanks

Former OpenAI staffers warn xAI safety failures could complicate SpaceX IPO

Its safety record is an “unpriced risk” that could complicate SpaceX’s plan to raise up to $75 billion in its expected IPO.

Defused News Writer profile image
by Defused News Writer
A rocket launches from a launch pad, surrounded by billowing clouds of smoke and steam. The image features a prominent logo for xAI in the foreground.

Former OpenAI staffers and several AI safety groups warned that xAI’s safety practices could create liabilities for SpaceX as the company prepares a large IPO.

The letter to investors flags “unpriced risks” tied to xAI and notes SpaceX’s private valuation rose to over $1 trillion after acquiring the AI lab, with plans to raise up to $75 billion in the offering.

The letter was published on Tuesday and reported by WIRED; its signatories include a new nonprofit, Guidelight AI Standards, cofounded by former OpenAI safety researcher Steven Adler and former OpenAI policy adviser Page Hedley, alongside Legal Advocates for Safe Science and Technology, Encode AI, and The Midas Project.

The authors say xAI’s record could increase regulation and litigation risk for the combined company and recommend disclosures on whether xAI will continue developing frontier AI models and a public safety and governance plan.

The letter also cites a recent SpaceX agreement to sell significant GPU capacity to Anthropic and says that deal leaves unclear whether xAI remains a frontier competitor inside SpaceX.

As examples of safety lapses the authors point to incidents where xAI’s chatbot Grok generated responses invoking white genocide and produced thousands of sexualized images of women and children that spread on X, prompting at least 37 US attorneys general to demand remedies.

The letter cites reporting that xAI had only “two or three” people working on safety as of January and argues managing frontier-AI risks requires substantial investment.

The authors acknowledge some recent steps, including expanded model testing with the White House, but call for further disclosures so investors can assess risk.

“xAI’s historical record has been serious enough to warrant scrutiny; it has not, by itself, foreclosed a better future for the company,” the letter reads.

Defused News Writer profile image
by Defused News Writer

Explore stories